Comments (5)
Do you have an example test case where the two scripts yield different results?
from cmu-thesis.
from cmu-thesis.
Thank you Yun for your quick reply! Included are 2 pickle files of pred + truth file of (20123,527) shape
from cmu-thesis.
Your eval script's AUC and dprime is about 50% of the sklearn one below:
import numpy as np
from scipy import stats
from sklearn import metrics
import torch
def d_prime(auc):
standard_normal = stats.norm()
d_prime = standard_normal.ppf(auc) * np.sqrt(2.0)
return d_prime
def calculate_stats(output, target):
"""Calculate statistics including mAP, AUC, etc.
Args:
output: 2d array, (samples_num, classes_num)
target: 2d array, (samples_num, classes_num)
Returns:
stats: list of statistic of each class.
"""
classes_num = target.shape[-1]
stats = []
# Accuracy, only used for single-label classification such as esc-50, not for multiple label one such as AudioSet
acc = metrics.accuracy_score(np.argmax(target, 1), np.argmax(output, 1))
# Class-wise statistics
for k in range(classes_num):
# Average precision
avg_precision = metrics.average_precision_score(
target[:, k], output[:, k], average=None)
# AUC
auc = metrics.roc_auc_score(target[:, k], output[:, k], average=None)
# Precisions, recalls
(precisions, recalls, thresholds) = metrics.precision_recall_curve(
target[:, k], output[:, k])
# FPR, TPR
(fpr, tpr, thresholds) = metrics.roc_curve(target[:, k], output[:, k])
save_every_steps = 1000 # Sample statistics to reduce size
dict = {'precisions': precisions[0::save_every_steps],
'recalls': recalls[0::save_every_steps],
'AP': avg_precision,
'fpr': fpr[0::save_every_steps],
'fnr': 1. - tpr[0::save_every_steps],
'auc': auc,
# note acc is not class-wise, this is just to keep consistent with other metrics
'acc': acc
}
stats.append(dict)
return stats
from cmu-thesis.
I've found the cause: the pred
array you passed in has the float16
dtype, and my function roc
isn't robust against it.
Below is what happened in detail:
def roc(pred, truth): # here, pred is float16 and truth is bool
data = numpy.array(sorted(zip(pred, truth), reverse = True)) # here, data is float16
pred, truth = data[:,0], data[:,1] # here, both pred and truth becomes float16
# now we're computing the cumsum on a float16 array.
# float16 doesn't have enough precision to represent 2049:
# when it adds 1 to 2048, the result is still 2048.
# therefore the cumsum is capped at 2048, and messed up everything from here.
TP = truth.cumsum()
FP = (1 - truth).cumsum()
mask = numpy.concatenate([numpy.diff(pred) < 0, numpy.array([True])])
TP = numpy.concatenate([numpy.array([0]), TP[mask]])
FP = numpy.concatenate([numpy.array([0]), FP[mask]])
return TP, FP
I've modified the second line of the function to convert truth
to bool
, and this fixes the problem.
from cmu-thesis.
Related Issues (7)
- download failed HOT 2
- SoundNet Transfer Code HOT 2
- predict 1s audio HOT 10
- Data processing HOT 2
- Gets Forbidden message while trying to download pretrained model HOT 5
- The data cannot download HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from cmu-thesis.